In this paper, a robust optimization design method is proposed and investigated, where a rare-earth-reduced high-torque-density permanent magnet (RH-PM) motor is selected as a research object. The key of this study is to introduce a new concept of parameter sensitivity region (PSR) into the process of optimization, aiming at comprehensively considering the fluctuation of multiple motor parameters into the motor design. To determine the motor parameters related to the sensitivity region, a parameters anti-disturbance analysis is carried out based on analysis of variance. According to the analysis, motor design parameters are classified into two groups, where the PSR-based uncertainty optimization method is purposely utilized in the group with highly robustness-related parameters. Then, a deterministic optimization is employed for the group with weakly robustness-related parameters. After optimization, the motor performances are analyzed in detail, including torque, torque ripple, and so on, as well as their corresponding analysis of performance consistency. Finally, a prototype is manufactured and tested to further verify the proposed method and RH-PM motor.